-
Notifications
You must be signed in to change notification settings - Fork 143
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Graded spike payload in pre-synaptic trace updates - floating-pt
and bit-approximate-loihi
#607
Conversation
…nc/lava into dev/learning_graded_spikes
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I have not reviewed this to a point, where I could approve or block it, so I am just leaving comments.
It all looks good though, at least from a high level coding-style point of view. I only added a few minor remarks.
The one thing we could maybe improve are the known-value tests. It is not clear where the known values come from or how we know they are correct. There are also so many that the significance of individual numbers is unclear. Please keep in mind that someone will at some point look at this code and need to understand why all these values as deemed correct. Make sure that it is very easy to understand that from the code. That person could be yourself one year from now. :)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This concept of how the payload modifies the traces seems quite complicated. We should have a good example in the learning tutorial to show what will happen.
If there is no issue for that yet, please create one.
…nc/lava into dev/learning_graded_spikes
…pikes # Conflicts: # src/lava/magma/core/model/py/connection.py
…nc/lava into dev/learning_graded_spikes
Issue Number: #606
Objective of pull request: As a user of the learning API, I would like to use graded spike payloads to update pre-synaptic traces in order to implement more complex learning algorithms. Especially, this is critical for prototype-based learning approaches.
Pull request checklist
Your PR fulfills the following requirements:
flakeheaven lint src/lava tests/
) and (bandit -r src/lava/.
) pass locallypytest
) passes locallyPull request type
Please check your PR type:
What is the current behavior?
What is the new behavior?
Does this introduce a breaking change?
Supplemental information